Hello all, i have a hard time finding the number of households with married couples that live with a roommate. I have used “related” and “marst” but the numbers don t seem to add up. I was trying to replicate the number & percentages from an article where it says that approx. 4.2 million households lived with a roomate (CPS) and for married couples the rate was just 0.46% (just over 280,000 households). The numbers at national level are from CPS and for metropolitan areas are from ACS. I would like to use only one source - mainly - ACS.
Thank you.
I would suggest identifying married couple households using RELATE. For households where the head is married and their spouse is in the household, there will be an individual with RELATE = 2. Once you’ve identified these households, you can identify the subset with roommates by the household having a member with RELATED = 1115. Once you’ve marked the relevant households, you should calculate household statistics by keeping only one household member (usually the head) and using HHWT as the weight.
Thank you for the quick answer. Is this operation possible using Online Data Analysis System or only using statistical software? Kind of a newbie in statistical software.
Thank you and have wonderful day!
Unfortunately I don’t think there is a practical way to do this without using a statistical software package. If you want to use a package and would like assistance please email ipums@umn.edu